House Speaker Mike Johnson (R-La.), who will attend a conference of artificial intelligence officials in Washington later Tuesday, said on CNBC’s “Squawk Box” that he supports “striking the right balance” with artificial intelligence technology and wants the guardrails to be “voluntary.”
“I hope it’s voluntary,” Johnson told CNBC on Tuesday before meeting with executives and President Donald Trump. “We’ll see what comes out in today’s meeting. I think they’re heading there. I think they see the handwriting on the wall. I think they want to assure investors, you know, consumers, the American people, that they’re going to do this in a very safe way. And I think that puts us in the right position.”
Johnson’s comments come amid pressure from industry executives and lawmakers to act quickly on legislation to establish a regulatory structure for AI and impose government restrictions aimed at preventing harm from AI. Speakers will meet with executives including Anthropic’s Dario Amodei, Meta’s Mark Zuckerberg and Alphabet’s Sundar Pichai.
Congress is still in the early stages of regulating artificial intelligence, and the Legislature is scheduled to adjourn from Washington this weekend until after the November election.
Sen. Mark Warner, D-Virginia, said his time away from Washington gives the issue a sense of urgency. “It’s completely unacceptable – I have this week off after this happens,” he told the Squawk Box show.
Warner, who made his fortune as a technology entrepreneur, said he is not an AI “doomer” but believes there are real risks rogue agents could pose to national security and critical infrastructure. On Tuesday, he plans to seek unanimous consent for a bill that would create a federal AI safety commission to review Frontier models at least 45 days before they are released to the public. This is a quick way to get widely popular legislation passed on the Senate floor. This bill is unlikely to pass.
AI executives recently warned that the technology they built poses catastrophic risks if it goes wrong, and called on the U.S. government to take action to slow advances in cutting-edge technology. But AI will support the economy and drive market gains amid other headwinds as the technology promises increased productivity and new breakthroughs.
OpenAI, whose developer day is Tuesday, announced late Monday that it has decided not to release its next artificial intelligence model, GPT-6.1 Astra, because it has determined that it does not fully meet the company’s safety standards.
Proposals regarding AI safety have been floating around Capitol Hill in recent weeks. Anthropic warned in its IPO prospectus that its technology poses a “catastrophic or even existential risk to humanity,” and expectations for regulation will continue to rise.
The regulatory proposals include establishing an independent auditor within Frontier Labs, introducing new liability laws and launching a regulatory review of AI models.
Asked whether the AI companies he met had raised the possibility of liability immunity in potential regulations, Johnson said: “Nothing like that has come up.”
“Of course, product liability law applies directly here, and we’re going to make sure that keeps everyone in check,” he said.
